the plea
homewatch onlinefaqsinterviewsdiscussion

personal failure, inistitutional failure, and the sixth amendment by albert w. alschuler
[From New York University Review of Law and Social Change, Volume XIV, Number 1, 1986, pages 149-156; Republished with the author's permission.]

I agree with Professor Schulhofer's main thesis: The problem of guaranteeing the effective assistance of counsel cannot be solved within a plea bargaining system.1 The reasons for this conclusion fall under two main headings. First, a system of plea negotiation is a catalyst for inadequate representation. It subjects defense attorneys to serious temptations to disregard their clients' interests, engenders suspicion of betrayal on the part of defendants, and aggravates the harmful impact of inadequate representation when it occurs. Second, a plea negotiation system insulates attorneys from review and often makes it impossible to determine whether inadequate representation has occurred.

Alschuler is the Professor of Law, Russell Baker Scholar and Acting Director, Center for Studies of Criminal Justice, University of Chicago; Member, Illinois Bar. A.B., Harvard University, 1962; LL.B., Harvard Law School, 1965.

As Professor Schulhofer suggests, almost every defense of plea negotiation depends on the assumption that defendants will be well-represented.2 Apologists for plea bargaining draw pictures of well-informed defendants, advised by capable attorneys, making rational assessments of surrender and gain.3 These apologists know that, were they to peer into the pit, they often would find their assumptions unjustified. Nevertheless, they regard the defective performance of lawyers as exogenous to a plea bargaining system. In their view, inadequate lawyers are as likely to appear and work their mischief in one system as another; whether their cases are bargained or tried is immaterial.4 Although these observers condemn departures from their idealized models as abuses, they evaluate the institution of plea bargaining by examining how the process might work with Earl Warren as the prosecutor, Socrates as the defense attorney, and Solomon as the trial judge. These observers wear blinders. They are somewhat like the people who once proclaimed that monarchy is a marvelous form of government so long as the king is good. Although apologists for plea negotiation usually recognize that the king may not be good, they argue that no system can be better than the people who administer it.


My principal goal in this paper is to demonstrate that institutional arrangements make a difference and that the performance of lawyers is not exogenous. A regime of plea negotiation lends itself to ineffective assistance and aggravates its harmful consequences in at least six ways that an adjudicative system5 does not.6

First, like other people, defense attorneys like money. A plea bargaining system subjects these attorneys to powerful financial temptations to disregard their clients' interests. For obvious practical reasons, privately retained defense attorneys usually collect their fees in advance; and once an attorney has pocketed his fee, his economic interests lie in disposing of the case as rapidly as possible. The most rapid way to dispose of a case is usually to enter a bargained plea. A similar contlict of interest besets some appointed attorneys -- those who receive small statutory payments for every case in which they appear. The statutory compensation is never a great reward for conducting a two-day trial; but for many lawyers, the burdens of representing the indigent become tolerable when these burdens consist of a brief conference with the client, a brief conference with the prosecutor, and a brief courtroom colloquy before a negotiated plea is accepted.7

Second, apart from the desire to make money, attorneys like to minimize work. Plea negotiation offers defense attorneys a more comfortable way of life than does adjudication. As salaried lawyers whose compensation does not depend on the ways in which their cases are resolved, public defenders are not subject to the same economic temptations as private defense attorneys. Nevertheless, all lawyers are subject to the temptation to promote their own convenience at the expense of their clients. As with the economic temptations noted above, the danger is not so much deliberate betrayal as warped judgment and excessive use of shortcuts. As one judge noted, "It is easier to sit in an overstuffed chair drinking coffee than to stand in the courtroom trying cases."8 It is also easier to banter with prosecutors about kids, lakehouses and justice than to march into the field to learn the facts or into the library to learn the law.

Somewhat similar conflicts of interest would influence the performance of defense attorneys in an adjudicative system. There are a variety of ways to bluff, cut corners, and "wing it" when a case is tried. Nevertheless, the trial process constrains the ability of lawyers to shrug off most forms of hard work as unnecessary. At trial, an attorney must at least listen to the prosecution's witnesses and decide whether and how to cross-examine them. Even if the attorney has conducted no pretrial investigation, his client may have suggested some defense witnesses (including the client himself). The attorney must either call these witnesses to testify or explain to his client why he has not. In addition, the attorney must participate in the formulation of the court's instructions, argue to the jury, and more. An attorney who wished to appear competent in performing these visible tasks would be likely to investigate the facts before trial, speak to his witnesses, research the relevant law and prepare in other ways. He would not have an easy way to conclude his representation in minutes and rationalize his lack of vigorous advocacy as the best possible representation of his client.

In a typical shop or factory, some employees work harder than others, but the possibility of defective performance on the job does not suggest that it is immaterial whether the manager of the shop or factory requires its employees to report for work. Similarly, the temptations to cut corners that defense attorneys would experience in an adjudicative system cannot be equated with the more powerful conflicts of interest that confront them in our plea bargaining system.

Third, like other people, attorneys like to be liked and to enjoy good relationships with co-workers. This personal interest, like the others, can lead defense attorneys to represent their clients less vigorously. In a plea bargaining system, prosecutors and trial judges -- the group with whom a defense attorney works every day -- are likely to become a more important constituency than the attorney's more transient clients.9 As with the other temptations that I have mentioned, scholars could write treatises about this one, and some organizational theorists have.10 What these theorists sometimes have been slow to recognize, however, is that there is a partial cure for the disease. An important virtue of the administration of justice by juries and other impartial tribunals is that it maximizes the extent to which outcomes depend on the facts of each case and minimizes the influence of personal favoritism and favor-seeking. Juries in particular are not repeat players. An adjudicative system treats each case as a tale in which the defendant has a major role rather than as a short chapter in which he has a minor part. It reduces the chance that the defendant's interests will be sacrificed to build capital for the future, repay old debts, or simply win the approval of co-workers for its own sake.

Plea bargaining promotes inadequate representation for a fourth reason. Defense attorneys, like other people, do not want to be proven wrong. A decision to plead guilty, unlike a decision to stand trial, cannot be proven wrong. A guilty plea not only masks prior errors and professional deficiencies but also ensures the impregnability of a lawyer's professional judgment. Once a guilty plea has been entered, no one can know what result a trial would have reached. From the defendant's perspective, one can always suppose that the outcome would have been worse. When an attorney takes a case to trial, by contrast, he knows (and his client usually knows) what offer or bargaining opportunity has been declined. At the conclusion of a trial, it may be evident that the rejection of an offer has cost a defendant several years of his life. This moment of recognition is unlikely to enhance a lawyer's self-esteem or make future contact with the client more pleasant. Nevertheless, the risk of this unhappy moment can be avoided. A lawyer need only follow what is always the safe and secure course in a plea bargaining system -- persuade the client to plead guilty.

A fifth reason why a regime of plea bargaining promotes inadequate representation is its secrecy and unwritten rules. These characteristics maximize the dangers of inexperience. As one prosecutor observed, "Anyone who can try a civil case can try a criminal case, but a civil lawyer is not qualified to evaluate a criminal case. He has no way of knowing what a criminal case is worth."11 Moreover, the danger is not simply that an inexperienced lawyer may be unable to distinguish a good offer from a bad offer. In addition, the lawyer may not know how to get a good offer. When does a lawyer push too hard and harm his clients because courthouse insiders conclude that he is a "shotgun who goes off half-cocked"? When does he fail to push hard enough and harm his clients because insiders realize that they can take advantage of him? How "reasonable" is too reasonable, and how "reasonable" is not reasonable enough? These questions are difficult, and defendants may suffer while attorneys try to figure them out.

By contrast, trial is an open system -- a system with reasonably well-defined roles and a system that can be studied, learned, and eventually mastered. Moreover, even when an inexperienced attorney founders at trial, the witnesses usually tell their stories in a coherent fashion; the judge usually gives reasonably accurate instructions on the law; the jurors usually follow both the evidence and their consciences; and if the defendant is convicted, the judge imposes the sentence that he thinks fair. An inexperienced defense lawyer can harm his clients greatly at trial, but to a far greater extent than plea negotiation, trial is a system of checks and balances.

There is a sixth way in which the performance of defense attorneys in a plea bargaining system differs from their performance in an adjudicative system. A conscientious lawyer in a plea bargaining system cannot be only an advocate for his client. The lawyer must also be the point man or woman for a coercive system of justice. It is the defense attorney who must deliver the message that the client does not have an unfettered right to trial. It is the defense attorney who must explain how the plea bargaining leverage works. Some clients may be slow to get the message, and one public defender observed, "A lawyer shirks his duty when he does not coerce his client."12 The duty of a conscientious lawyer is indeed to ensure that his client understands what is at stake. In performing this duty, the lawyer may be required to use such harsh language as: "I cannot beat this case. The jury wants your blood. You will burn if you do not change your plea."13

Although a well-intentioned lawyer who "coerces" his client may be effective in saving the client from an avoidable penalty, he is likely to be ineffective in another respect. He is unlikely to inspire his client's confidence. The client may conclude that the lawyer does not believe in his case. In a plea bargaining system, even a capable and conscientious defense attorney is likely to incur the distrust and resentment of the person whom he seeks to serve. One can envision a different role for a defense attorney -- a role in which the attorney could leave the judging to others and be unreservedly on his client's side. Nevertheless, an attorney could properly assume this role only in a system that did not penalize the decision to present an adversary defense.

Perhaps, for some lawyers, the lofty ideals of our profession are an adequate answer to the temptations that I have described; but I confess that, were I a defense attorney, these temptations would influence me. I might not join the ranks of those lawyers who purposely betray their clients, who appear in hundreds of cases every year and never try any of them, and who sometimes even deceive their clients in order to persuade them to plead guilty. Nevertheless, it is a rare case in which a defense attorney cannot secure a plausible offer from a prosecutor. After receiving a colorable offer, I could not forget that the alternative to accepting it would be extra days of work -- work which might or might not benefit my client but which certainly would prove costly to me.


I will address only briefty the second set of reasons why insuring the effective assistance of counsel is and will remain a dream in our plea bargaining system. In a system whose structure encourages inadequate assistance, no effective mechanism exists for determining whether inadequate assistance has occurred. Nor does it seem likely that an effective mechanism can be created.

In 1984, after years of neglecting the issue, the Supreme Court articulated a general standard for judging the effectiveness of counsel. In Strickland v. Washington14, a defendant who had pleaded guilty challenged the effectiveness of his counsel at the sentencing proceeding that followed his plea. The Supreme Court viewed the case as a vehicle for addressing in sweeping legislative fashion a broad range of issues likely to arise in effective assistance cases.15 The Court said:

In giving meaning to the [Constitutional] requirement, . . . we must take its purpose -- to ensure a fair trial -- as the guide. The benchmark for judging any claim of ineffectiveness must be whether counsel's conduct so undermined the proper functioning of the adversarial process that the trial cannot be relied on as having produced a just result.16

In a system in which almost no one receives a trial, the terms of the Court's long-awaited general formula for resolving claims of ineffective assistance were limited to trials. Perhaps this limitation reftected the myopic assumption that an adversarial trial is still the norm in the American system of criminal justice. The limitation, however, may have reflected, not the Supreme Court's naivety, but its sophistication. Only a defendant who has exercised the right to trial has a significant chance of demonstrating that he has not received the effective assistance of counsel. Perhaps, for this reason, the Supreme Court concluded that guilty plea cases could be safely disregarded.

Despite the formal limitation of the Strickland standard to trials, the proper application of the Court's approach to claims of ineffective assistance in the guilty plea process seems apparent. The Strickland Court said:

First, the defendant must show that counsel's performance was deficient. This requires showing that counsel made errors so serious that counsel was not functioning as the "counsel" guaranteed by the Sixth Amendment. Second, the defendant must show that the deficient performance prejudiced the defense. This requires showing that counsel's errors were so serious as to deprive the defendant of a fair trial, a trial whose result is reliable.17

The court added that "[j]udicial scrutiny of counsel's performance must be highly deferential," that "a court must indulge a strong presumption that counsel's conduct falls within the wide range of reasonable professional assistance," and that "[t]he defendant must show that there is a reasonable probability that, but for counsel's unprofessional errors, the result of the proceeding would have been different."18

Strickland imposes a heavy burden on defendants convicted at trial who challenge the adequacy of the representation that they have received. At a trial, however, both the attorney's courtroom performance and the evidence presented by the prosecutor are matters of record. On rare occasion, a defendant may be able to point to specific misconduct, demonstrate that it was egregious, and show that competent representation probably would have altered the outcome. In a guilty plea case, by contrast, both the attorney's conferences with his client and his conferences with the prosecutor are secret. The absence of a record makes it extraordinarily difficult for a defendant to demonstrate that his plea was unreliable or that counsel's errors were likely to have altered the outcome. Moreover, it is difficult for a reviewing court to know what can be expected of an attorney in an unstructured bargaining situation. As Professor Schulhofer has noted, one can conjure up a plausible reason for almost any default in a plea bargaining system.19

Strickland's heavy weighing of the scales against claims of ineffective assistance drives the last nail into the coffin of the defendant whose inadequate lawyer has led him to plead guilty. But this last nail hardly matters. Whatever the formula for judging claims of ineffective assistance and whatever the facts, the defendant who has been inadequately represented and induced to plead guilty has almost no chance.

Ineffective representation is far more likely to occur in guilty plea cases than in trial cases. The weakest members of the defense bar are not usually the lawyers who put the state to its proof. Committing serious errors at trial might be an enormous step forward for some attorneys. Nevertheless, defendants who have pleaded guilty are far less likely to secure relief on grounds of ineffective representation than defendants who have been convicted at trial.

This irony is compounded by the fact that, in the Brady trilogy in 197020 the Supreme Court equated a knowing waiver with a competently counseled waiver. The Court therefore refused to consider directly whether the defendants in the three cases before it had knowingly waived their constitutional rights by entering gnilty pleas. With rare exceptions, the Court has since permitted defendants who have pleaded guilty to secure relief only by shouldering the all-but-impossible burden of demonstrating that they did not receive the effective assistance of counsel.21

The American system of criminal justice is distinctive in three respects. First, it makes the kind of justice that a defendant receives more dependent on the quality of counsel than any other legal system in the world. Second, it subjects defense attorneys to serious temptations to disregard their clients' interests. And third, it makes it impossible to determine whether defendants have received the effective assistance of counsel. No proposed reform except Professor Schulhofer's bold new model -- the adversary criminal trial -- holds promise of eliminating these paradoxes.

* EDS. NOTE: Normally it is the policy of the Review to use female pronouns for the third person singular when the pronoun is used generically. However, at the time this paper was accepted for publication, this policy was not mandatory.


1 Schulhofer, Effective Assistance on the Assembly Line, 14 N.Y.U. REV. L. & SOC. CHANGE 137 (1986) [hereinafter cited as Schulhofer, Effective Assistance].

2 Cf., e.g., id. at 140, 144.

3 E.g., Brady v. United States, 397 U.S. 742 (1970); Easterbrook, Criminal Procedure as a Market System, 12 J. LEGAL. STUD. 289 (1983); Church, In Defense of "Bargain Justice", 13 LAW & SOC'Y REV. 509 (1979); Hyman, Bargaining and Criminal Justice, 33 RUTGERS L. REV. 3 (1980).

4 E.g., Easterbrook, supra note 3, at 309-10; Church, supra note 3, at 521; Hyman, supra note 3, at 14.

5 I use the term "adjudicative system" to include every system that does not reward or encourage waiver of the right to trial. The term is not limited to the sorts of systems found in Continental European nations -- systems in which, with minor cases set aside, the guilty plea is unknown so that every case is tried.

6 My contention is not that all or most defendants fail to receive effective assistance in a plea bargaining system. It is that inadequate representation is likely to occur in a significant number of cases, that it cannot be remedied effectively, and that the problem would be alleviated by moving toward an adjudicative system of the sort advocated by Professor Schulhofer. Similarly, I do not claim that the deficiencies of a plea bargaining system discussed in this commentary would, without more, warrant the abandonment of plea bargaining. These remarks merely explore some of the many subordinate injustices to which a fundamentally unjust practice has led.

7 Although a defense attorney's financial interests usually favor the entry of a plea of guilty, these interests occasionally push in the opposite direction. Consider, for example, an assigned counsel system that compensates attorneys on an hourly basis. Attorneys who regard the prescribed hourly rate as inadequate may be tempted to put the burdens of their assignments behind them by entering bargained pleas of guilty. Attorneys who lack more remunerative uses of their time, by contrast, may be tempted to "milk" their assignments. These attorneys may take cases to trial when that course would not be in their clients' interests. The point is simply that a defense attorney often has personal interests in his client's choice of plea that differ from those of the client himself.

8 See Alschuler, The Trial Judge's Role in Plea Bargaining, 76 COLUM. L. REV. 1059, 1103 (1976) (quoting Judge Arthur L. Alarcon).

9 For an in-depth exploration of this phenomenon, see Guggenheim, Musings on Some Ethical Dilemmas for the Institutional Criminal Defense Attorney, 14 N.Y.U. REV. L. & SOC. CHANGE 13 (1986).


11 Alschuler, The Defense Attomey's Role in Plea Bargaining, 84 YALE L.J. 1179, 1269 (1975).

12 Id. at 1310.

13 Cf. Huot v. Commonwealth, 363 Mass. 91, 94, 292 N.E.2d 700, 702 (1973) (language similar to that quoted in text).

14 466 U.S. 668 (1984).

15 Cf. Miranda v. Arizona, 384 U.S. 436 (1966). The Miranda style of opinion writing, in which a court roams at large through an area of the law, promulgates a doctrinal code, and treats the facts before it almost as an afterthought, was unusual at the time of Miranda itself. Thanks to a remaking of the Supreme Court in the name of "strict construction," however, this nonjudicial style now appears to be the norm.

16 Strickland, 466 U.S. at 686.

17 Id.

18 Id. at 687, 689, 694.

19 Schulhofer, Effective Assistance supra note 1, at 142.

20 Brady v. United States, 397 U.S. 742 (1970); McMann v. Richardson, 397 U.S. 759 (1970); Parker v. North Carolina, 397 U.S. 790 (1970).

21 Compare Tollett v. Henderson, 411 U.S. 258 (1973) with Blackledge v. Perry, 417 U.S. 21 (1974) and Menna v. New York, 423 U.S. 61 (1975).


home · introduction · case: stewart · case: gampero · case: jarrett · case: cook · faqs · interviews
discussion · producer's chat · readings & links · press reaction · tapes & transcripts · credits
privacy policy · FRONTLINE home · wgbh · pbsi

posted june 17, 2004

FRONTLINE is a registered trademark of wgbh educational foundation.
shadow photo copyright © michael s. yamashita/corbis
web site copyright 1995-2014 WGBH educational foundation